Adam Bede

    Dead Internet Theory

    Subtitle

    Date
    Tags
    Type

    The Title

    "The Internet Is Dead, Long Live the Internet" borrows the monarchical succession formula ("The King is dead, long live the King") — which itself is a statement about continuity through rupture. The old king dies; the new king is already king. No gap. That's the structural tension your essay lives in: the internet we knew is gone, something has replaced it, and the replacement happened without a clean funeral.

    The Argument Across Legitimacy Levels

    The piece would work as a ladder — starting with the most conspiratorial version of "Dead Internet Theory" and climbing toward the most defensible, showing that each rung has something worth taking seriously even as you shed the tinfoil:

    Level 1: The Conspiracy (entertaining but wrong)

    The actual Dead Internet Theory from ~2021: most online content is bot-generated, most "people" are AI, and the whole thing is a psyop. This is where your senior thesis aside lands — you can wink at it. It's wrong as a literal claim. But it feels true, and that feeling is doing real work. Why?

    Level 2: The Vibes Are Right (experientially true)

    The internet feels dead because the thing that made it alive — serendipity, human weirdness, stumbling onto something no algorithm served you — has been systematically optimized away. Kyle Chayka's Filterworld is your anchor here. Recommendation engines homogenize taste. Everything looks the same. The "cozy web" retreats into group chats and Discord servers. The public internet becomes a performative wasteland. It's not dead — it's undead. Zombie internet.

    Level 3: The Economics (structurally true)

    Cory Doctorow's enshittification thesis. Platforms follow a lifecycle: attract users → attract businesses → extract value from both → decay. The internet isn't dead — it's been strip-mined. This is the most defensible structural argument. The open web that let your senior thesis get 500+ downloads is not the same internet that now exists behind paywalls, login walls, and algorithmic gates.

    Level 4: The Epistemological Problem (philosophically true — and this is where it connects to R² Metaphors)

    If LLMs are trained on the internet, and LLM-generated content floods back into the internet, and the next generation of LLMs trains on that — you get model collapse. The internet becomes a recursive mirror. Your "bed of leaves" 🍂 problem. The internet isn't dead — it's incestuous. The map is training the next map, and no one's checking against the territory anymore. This is where Simon's satisficing becomes dangerous: we've satisficed our way into an information ecosystem where "good enough" outputs are indistinguishable from human ones, and the compounding imprecision is invisible.

    Level 5: The Succession (the "long live" part)

    So what's the new king? This is where you get to make your move. Options:

    • The internet is dead → long live the curated internet (human editorial judgment as the premium)
    • The internet is dead → long live the local internet (community, trust networks, the cozy web)
    • The internet is dead → long live the verified internet (provenance, watermarking, proof-of-human)
    • Or, more interestingly: the internet is dead → and we don't know what the new king looks like yet, and that's the point. The succession formula assumes continuity. What if this time, there's actually a gap?

    Tone & Voice

    Given your existing essays, I'd lean into the same structure: punchy intro with emoji energy, a few data bullets that earn the argument, then the longer-form philosophical escalation. The ladder from conspiracy → vibes → economics → epistemology → succession gives you natural section breaks. And the self-deprecating thesis aside ("Proving the internet is dead 💀 b/c no one wants to read that") becomes the emotional through-line — you wrote about this before anyone cared, and now the conspiracy theorists accidentally stumbled into something real.

    Key Thinkers to Draw On

    • Chayka — Filterworld, algorithmic homogenization
    • Doctorow — enshittification, platform decay
    • Shumailov et al. — model collapse (the technical paper on recursive AI training)
    • Jaron Lanier — You Are Not a Gadget, the original warning about what gets lost when we flatten human expression into templates
    • Hossein Derakhshan — the Iranian blogger who went to prison for 6 years and came back to find the web had died while he was away (this is an incredible essay and would make a killer epigraph)
    • Simon (already in your toolkit) — satisficing as the mechanism by which we accept the degradation
    • O'Gieblyn (already in your toolkit) — bidirectionality applies directly: we built the internet to reflect us, and now it's reshaping what "us" means